Abstract
Liver hepatocellular carcinoma (LIHC) is an aggressive cancer associated with chronic liver disease, necessitating better biomarkers and therapies. Manganese, an essential trace element, regulates tumor development. Data from TCGA and GEO databases were analyzed to identify manganese metabolism-related genes (MMRGs). LIHC samples were classified into subtypes via consensus clustering. A prognostic model was developed using LASSO and multivariate Cox regression, then validated using ROC and survival analysis. Immune infiltration was assessed via ssGSEA and CIBERSORT, and cell communication was analyzed with single-cell data (GSE149614). Tumor Mutational Burden (TMB), drug sensitivity, and a nomogram were also evaluated. Two manganese metabolism-related subtypes were identified, with Cluster 1 showing better survival. A two-gene model (CEP55 and SPP1) reliably predicted poor prognosis in high-risk groups. The high-risk group exhibited distinct immune profiles, including increased immune infiltration, elevated checkpoint expression, and higher TIDE scores. Single-cell analysis revealed altered T cell communication. This study established manganese metabolism-related subtypes and a prognostic model for LIHC, providing insights into immunoregulation and cell communication to guide precision diagnosis and immunotherapy.